Synthesis of tetrahydrofurans by a tandem hydrogen atom abstraction/ radical nucleophilic displacement sequence

Article abstract of DOI:10.1021/ol990564b

(equation presented) The reaction of a series of 5-(N-phthalimidoxy)-1-phenyl-1-(diphenylphosphatoxy)pentanes with triphenyltin hydride and AIBN provides alkoxy radicals which undergo 1,5-hydrogen atom abstraction to give β-(phosphatoxy)alkyl radicals. These radicals then take part in a radical nucleophilic displacement leading, after chain transfer, to tetrahydrofurans.

Full text of DOI:10.1021/ol990564b

ORGANIC
LETTERS
1999
Vol. 1, No. 2
225-227
Synthesis of Tetrahydrofurans by a
Tandem Hydrogen Atom Abstraction/
Radical Nucleophilic Displacement
Sequence
,1b
David Crich,*,1a Xianhai Huang,1a and Martin Newcomb*
Department of Chemistry, UniVersity of Illinois at Chicago, 845 West Taylor Street,
Chicago, Illinois 60607-7061, and Department of Chemistry, Wayne State UniVersity,
Detroit, Michigan 48202
dcrich@uic.edu
Received April 8, 1999
ABSTRACT
The reaction of a series of 5-(N-phthalimidoxy)-1-phenyl-1-(diphenylphosphatoxy)pentanes with triphenyltin hydride and AIBN provides alkoxy
radicals which undergo 1,5-hydrogen atom abstraction to give â-(phosphatoxy)alkyl radicals. These radicals then take part in a radical nucleophilic
displacement leading, after chain transfer, to tetrahydrofurans.
Well-planned tandem radical sequences, enabling the con-
struction of complex molecular architectures from relatively
simple precursors, are some of the most powerful transforms
in modern organic chemistry.2 Radical polar crossover
sequences, in which a radical reaction is coupled to an
oxidation or reduction followed by a two-electron process,
represent a possible means of further augmenting the power
of tandem schemes. For example, in the reductive sphere,
samarium diiodide allows radical processes to be coupled
to subsequent nucleophilic processes, via the trapping of
radicals as organosamarium III species.3-5 On the other hand,
the potential of radical cyclizations followed by oxidation
to cations and eventual trapping has been elegantly demon-
strated by Murphy using the combination of arenediazonium
salts and tetrathiafulvalene.6,7
We recently described a novel radical process,8 predicted
computationally by Zipse,9,10 in which an alcohol displaced
a phosphate group from a â-(phosphatoxy)alkyl radical (1)
to give the vicinally substituted radical (2) (Scheme 1).
Preliminary kinetic8 and stereochemical11 data were con-
sistent with either a concerted displacement on the initial
radical or trapping of a contact ion pair arising from
heterolysis of the â-(phosphatoxy)alkyl radical. We set out
below our first attempts to develop this type of vicinal radical
nucleophilic substitution into a synthetically useful tandem
type sequence.
The synthesis of tetrahydrofurans presented focuses on the
cyclization onto the â-(phosphatoxy)alkyl radical with vicinal
displacement of the phosphate group. The initial challenge
(6) Bashir, N.; Callaghan, O.; Murphy, J. A.; Ravishanker, T.; Roome,
S. J. Tetrahedron Lett. 1997, 38, 6255.
(1) University of Illinois at Chicago; (b) Wayne State University.
(2) For a collection of reviews dealing with all aspects of tandem
reactions, see: Chem. ReV. 1996, 96, issue 1 (Wender, P. A., Ed.).
(3) Curran, D. P.; Fevig, T. L.; Jasperse, C. P.; Totleben, M. J. Synlett
1992, 943.
(4) Molander, G. A.; Harris, C. R. Tetrahedron 1998, 54, 3321.
(5) Molander, G. A.; Harris, C. R. Chem. ReV. 1996, 96, 307.
(7) Callaghan, O.; Lampard, C.; Kennedy, A. R.; Murphy, J. A.
Tetrahedron Lett. 1999, 40, 161.
(8) Choi, S.-Y.; Crich, D.; Horner, J. H.; Huang, X.; Martinez, F. N.;
Newcomb, M.; Wink, D. J.; Yao, Q. J. Am. Chem. Soc. 1998, 120, 211.
(9) Zipse, H. Angew. Chem., Int. Ed. 1994, 33, 1985.
(10) Zipse, H. J. Chem. Soc., Perkin Trans. 2 1997, 2691.
(11) Crich, D.; Gastaldi, S. Tetrahedron Lett. 1998, 39, 9377.
10.1021/ol990564b CCC: $18.00 © 1999 American Chemical Society
Published on Web 05/29/1999

acid. The 1H NMR spectrum of the reaction mixture revealed
the exclusive formation of 2-benzyltetrahydrofuran (12),
which was subsequently isolated in 95% yield (Table 1).
Scheme 1. Radical Nucleophilic Displacement
Table 1. Formation of Tetrahydrofurans17
was one of designing suitable precursors for the â-(phos-
phatoxy)alkyl radical that are compatible with an intramo-
lecular nucleophile. Thus, it was felt that PTOC esters, as
used in the initial kinetic work, or bromides, as employed
in the stereochemical analysis, would not be suitable owing
to the potential for nucleophilic cyclizations prior to radical
cyclization. Phenylthio or phenylselenyl groups were found
to be unsuitable precursors owing either to a lack of reactivity
toward tin hydrides or to instability with respect to thiiranium
or seleniranium ion formation. In light of these potential
complications, we elected to take advantage of the built-in
alcohol and to use it, in the form of an alkoxyl radical, to
generate the â-(phosphatoxy)alkyl radical by 1,5-hydrogen
atom abstraction. This approach to the problem, aside from
giving rise to a tandem radical/polar sequence (Scheme 2),
Scheme 2. Tandem Hydrogen Abstraction/Cyclization
Sequence
has the considerable advantage of simplifying the substrate
and its preparation. One obvious pitfall, which could only
reliably be addressed experimentally, was the potential for
a competing 1,6-hydrogen abstraction from the benzylic
position.
The tin hydride-mediated cleavage of N-alkoxyphthal-
imides12 was selected as an appropriate means of alkoxy
radical generation, and this functionality was readily incor-
porated into a number of phosphate esters (6-11) by
straightforward procedures as outlined in the Supporting
Information.
When the same reaction conditions were applied to the
analogous diethyl phosphate 7, the outcome was less clear-
1
cut. The H NMR spectrum indicated the formation of 12
and two other substances, which were subsequently assigned
as the product of â-(phosphatoxy)alkyl migration13-16 (16)
(Figure 1) and the reduction product (17), in the ratio 60/
25/15 (Table 1). The 31P NMR spectrum, with three signals,
was consistent with this interpretation. Evidently, the nature
Treatment of a solution of 6 in benzene at reflux dropwise
with Ph3SnH and AIBN resulted in a crude reaction mixture
containing a single resonance in its 31P NMR spectrum (δ
-23.1), which was assigned to a salt of diphenylphosphoric
(13) Crich, D.; Yao, Q. J. Am. Chem. Soc. 1993, 115, 1165.
(14) Crich, D.; Yao, Q.; Filzen, G. F. J. Am. Chem. Soc. 1995, 117,
11455.
(15) Choi, S.-Y.; Crich, D.; Horner, J. H.; Huang, X.; Newcomb, M.;
Whitted, P. O. Tetrahedron 1999, 59, 3317.
(16) Beckwith, A. L. J.; Crich, D.; Duggan, P. J.; Yao, Q. Chem. ReV.
1997, 97, 3273.
(12) Kim, S.; Lee, T. A.; Song, Y. Synlett 1998, 501.
226
Org. Lett., Vol. 1, No. 2, 1999

Self-evidently, the benzylic C-H bond is the weaker one,
and we are therefore led to attribute this regioselectivity to
a polar effect in which the strongly withdrawing phosphate
groups destabilize the transition state for benzylic hydrogen
abstraction by the electrophilic alkoxyl radical. Obviously,
the usual entropic factors also favor the 1,5-abstraction
observed. Second, the substituent effects, particularly the
contrast between 7 and 9, and lack of stereoselectivity (10
and 11) strongly suggest that the reaction is proceeding via
a stepwise fragmentation of the â-(phosphatoxy)alkyl radical
to give a styrene radical cation/phosphate anion pair (21)
followed by ring closure to give the more stable benzylic
radical (Scheme 3). Indeed, recent work on the mechanism
of the â-(phosphatoxy)alkyl rearrangement is strongly sup-
portive of a dissociative mechanism.18
Figure 1.
of the leaving groups has a significant effect on the outcome
of the reaction.
Scheme 3. Proposed Dissociative Mechanism
Application of the same conditions to the substrates 8 and
9, containing a methyl group at the site of hydrogen atom
abstraction, resulted in both instances in the formation of
the tetrahydrofuran 13 in excellent yield (Table 1). 31P NMR
spectroscopy of the crude reaction mixtures indicated only
one significant phosphorus-containing product in each case,
namely, the expelled phosphate. Thus, the inclusion of a
methyl group at the site of abstraction and substitution
facilitates displacement of the weaker leaving group diethyl
phosphate.
Moving the methyl substituent one and two carbons along
the aliphatic chain, as in 10 and 11, led to the isolation of
the tetrahydrofurans 14 and 15 in excellent yields (Table 1).
Unfortunately, there was no diastereoselectivity in either of
these two cyclizations under the conditions that were
employed. The remote possibility exists that the individual
diastereomers of both 10 and 11 cyclize with excellent but
opposite selectivity, resulting in the observed mixtures.
However, we consider this to be extremely unlikely in light
of the very poor selectivity noted in the intermolecular
displacements previously studied.11
All of the above examples result in the formation of
benzylically stabilized radicals. However we anticipate, from
consideration of the relative ease of â-(phosphatoxy)alkyl
rearrangements,14 that simple tertiary alkyl radical formation
would be sufficient.
Further studies on the mechanisms of these intriguing
reactions as well as on their applications in synthesis are
currently underway and will be described in due course.
Acknowledgment. We thank the NSF (Grants CHE
9625256 and CHE 9614968) for support to D.C. and M.N.,
respectively, and the University of Illinois at Chicago for a
University Fellowship (X.H.).
Supporting Information Available: Description of the
preparation of all substrates and characterization data. This
material is available free of charge via the Internet at
http://pubs.acs.org.
We also prepared the three acetates 18-20 but found no
evidence for tetrahydrofuran formation in any of their
reactions with Ph3SnH. These results are consistent with the
notion that acetate is an even weaker nucleofuge than diethyl
phosphate.
OL990564B
Two major points emerge from the above results. First,
1,6-hydrogen atom abstraction from the benzylic position
did not compete effectively with 1,5-hydrogen atom abstrac-
tion leading to the requisite â-(phosphatoxy)alkyl radicals.
(17) New compounds had spectral and microanalytical or HRMS data
consistent with the assigned structures.
(18) Whitted, P. O.; Horner, J. H.; Newcomb, M.; Huang, X.; Crich, D.
Org. Lett. 1999, 1, in press.
Org. Lett., Vol. 1, No. 2, 1999
227